| The purpose of this study was to determine test characteristics of the Pediatric Language Acquisition Screening Tool for Early Referral-Revised (PLASTER-R), a new tool for monitoring communication development in the pediatric population. Data from the PLASTER-R (Shulman & Sherman, 1996), a set of developmental questionnaires for children three to 60 months of age, were analyzed against data obtained from the Preschool Language Scale-3 (PLS-3) (Zimmerman, Steiner, & Pond, 1992) to determine specificity and sensitivity ratios, as well as to gather reliability and validity data on the PLASTER-R instrument. The PLASTER-R is divided into seven age screens, and each age screen contains five questions pertaining to receptive language/audition and five questions relating to expressive language. The two age screens for children between 37 and 60 months of age also contain a question pertaining to speech intelligibility. Parents completed two PLASTER-Rs within two weeks of each other. Specificity and sensitivity ratios were calculated for the first and second administrations of the PLASTER-R. Specificity and sensitivity ratios also were calculated using different PLASTER-R scoring criteria. Using the original scoring recommendations, the PLASTER-R was moderately successful in identifying children who were within normal limits for language development. However, when the scoring criteria were altered to take into account children's emerging language skills, specificity increased significantly. Results of the present investigation suggest the sensitivity of the PLASTER-R is poor. However, since only nine out of the 84 subjects in this study failed the PLS-3, this should not be considered conclusive evidence of poor sensitivity. It is suggested that future research be conducted with a greater number of subjects, including subjects with known language delays. The inclusion on the PLASTER-R of a question relating to speech intelligibility for children 37 months of age and older appeared to be beneficial. Overall, parental responses to this question were deemed more accurate than results from the PLS-3 Articulation Screener. Test-retest reliability for the PLASTER-R was high (r =.86). As with the specificity ratio, the concurrent validity of the PLASTER-R was significantly strengthened when subjects' emerging language skills were taken into account. |
